Add support for surrogates when serializing and deserializing for native ports

runtime/vm/dart_api_message.cc

Index: runtime/vm/dart_api_message.cc
diff --git a/runtime/vm/dart_api_message.cc b/runtime/vm/dart_api_message.cc
index 4cee392ee75564f19ef8bdd909d8116b9ed76e27..e22a54f03cf6d55308266441d80dc5c2fb9e036b 100644
--- a/runtime/vm/dart_api_message.cc
+++ b/runtime/vm/dart_api_message.cc
@@ -389,17 +389,21 @@ Dart_CObject* ApiMessageReader::ReadInternalVMObject(intptr_t class_id,
       uint16_t *utf16 =
           reinterpret_cast<uint16_t*>(::malloc(len * sizeof(uint16_t)));
       intptr_t utf8_len = 0;
+      // Read all the UTF-16 code units.
       for (intptr_t i = 0; i < len; i++) {
         utf16[i] = Read<uint16_t>();
-        // TODO(sgjesse): Check for surrogate pairs.
-        utf8_len += Utf8::Length(utf16[i]);
+      }
+      // Calculate the UTF-8 length.
+      Utf16::CodePointIterator it(utf16, len);
+      while (it.Next()) {
+        utf8_len += Utf8::Length(it.Current());
       }
       Dart_CObject* object = AllocateDartCObjectString(utf8_len);
       AddBackRef(object_id, object, kIsDeserialized);
       char* p = object->value.as_string;
-      for (intptr_t i = 0; i < len; i++) {
-        // TODO(sgjesse): Check for surrogate pairs.
-        p += Utf8::Encode(utf16[i], p);
+      it.Reset();
+      while (it.Next()) {
+        p += Utf8::Encode(it.Current(), p);
       }

[siva, 2012/11/28 22:02:24]

At this point here we could state that if malformed Dart strings are trying to
be passed out to the native world we could flag them as error. That way we can
say that only valid utf8 strings will be sent to the native world.

[Søren Gjesse, 2012/11/29 09:06:14]

Done.

       *p = '\0';
       ASSERT(p == (object->value.as_string + utf8_len));
@@ -807,7 +811,11 @@ bool ApiMessageWriter::WriteCObjectInlined(Dart_CObject* object,
       if (type == Utf8::kLatin1) {
         uint8_t* latin1_str =
             reinterpret_cast<uint8_t*>(::malloc(len * sizeof(uint8_t)));
-        Utf8::DecodeToLatin1(utf8_str, utf8_len, latin1_str, len);
+        bool success = Utf8::DecodeToLatin1(utf8_str,
+                                            utf8_len,
+                                            latin1_str,
+                                            len);
+        ASSERT(success);
         for (intptr_t i = 0; i < len; i++) {
           Write<uint8_t>(latin1_str[i]);
         }
@@ -816,7 +824,8 @@ bool ApiMessageWriter::WriteCObjectInlined(Dart_CObject* object,
         // TODO(sgjesse): Make sure surrogate pairs are handled.
         uint16_t* utf16_str =
             reinterpret_cast<uint16_t*>(::malloc(len * sizeof(uint16_t)));
-        Utf8::DecodeToUTF16(utf8_str, utf8_len, utf16_str, len);
+        bool success = Utf8::DecodeToUTF16(utf8_str, utf8_len, utf16_str, len);
+        ASSERT(success);
         for (intptr_t i = 0; i < len; i++) {
           Write<uint16_t>(utf16_str[i]);
         }